Two ways to mount a people counter
Most people counters end up in one of two places: bolted to the ceiling over the entrance, or fixed into the door frame itself. The picture looks similar from outside the store. The cost to install, the accuracy you get back, and what happens on a busy Saturday are not similar at all. This is a procurement piece for whoever owns the hardware decision: a head of retail operations, a facilities lead, a city analytics manager, or anyone signing off on the install bill for a people counting rollout across multiple sites.
The short version: a ceiling-mount counter looks down on the entrance and reads geometry from above. A door-frame counter sits at human height on the jamb of the doorway and reads horizontally across the threshold. Both can count well in the right room. Neither is universally better. The right pick depends on the entrance you have, the wiring you can run, and how much a missed count costs you.
Sight-line geometry: what each mount actually sees
Ceiling-mount: a top-down view of the threshold
A ceiling-mount sensor sits on the soffit, three to four metres above the entrance floor, and looks straight down. With a Time-of-Flight depth sensor, what it reads is a height map of whatever passes underneath. A walking adult is a tall blob moving at walking pace. A child is a shorter blob at a different pace. A shopping trolley with a child in it reads as two distinct heights stacked on the same footprint. Two adults walking shoulder to shoulder occupy two adjacent footprints under the lens, which the sensor sees as two people, not one.
Top-down geometry is the property that makes ceiling-mount the default for high-throughput entrances. Because the sensor is reading height, not silhouette, it does not confuse a parent and child for one person, and it does not lose a second visitor walking close behind a first. The depth map separates them in three dimensions.
Door-frame: a horizontal beam at the threshold
A door-frame counter sits on the side of the doorway at roughly chest or shoulder height. Most door-frame counters use a horizontal sensing beam, often a stereo pair or a beam-break, that triggers when something crosses the threshold from one side to the other. Some use a small camera-derived silhouette read from the side; others use simple infrared beam-break that just registers a crossing.
Horizontal sensing is fine on a narrow single-door entrance where people enter one at a time, in a line, with clear gaps between them. The geometry stops being kind the moment that pattern breaks: two adults walking abreast across a wide opening, a couple holding hands, a group entering as a cluster, a person pushing a pram, a wheelchair followed closely by a companion. From the side, those are hard to separate. From above, they are not.
Occlusion and group handling
Occlusion is the technical word for one person blocking another from the sensor's view. It is the single biggest source of miscounts in people counting, and it is where mount choice matters most.
- Ceiling-mount handles occlusion well. From above, two people walking side by side occupy two distinct footprints in the depth map. A small child next to a parent shows as a separate, shorter shape, not a piece of clothing on the adult. A group of four entering together resolves as four individual height blobs, because each person carries their own footprint into the field of view.
- Door-frame is more occlusion-sensitive. A horizontal beam or stereo pair sees the people closest to it. The second person walking just behind the first is partly hidden by the first body, and a side beam can read a couple walking close together as a single crossing. Group entries are the hardest case: a four-person family squeezing through together can register as one, two, or three depending on geometry and timing.
In a low-volume entrance where people queue up and enter singly, the difference does not show. In a busy retail or transit entrance, where groups, prams, trolleys, and side-by-side walking are constant, the difference shows on every minute of every busy hour.
Multi-doorway coverage
Real-world entrances often have more than one door panel. A double-leaf shopfront, a triple-door mall entrance, a wide hotel lobby with two side doors flanking a main one. The number of devices needed to cover the entrance depends on the mount style and on how wide each sensor can see.
Ceiling-mount across a wide entrance
A ceiling-mounted Time-of-Flight sensor at three to four metres typically covers a count zone roughly two and a half to four metres wide, depending on each sensor's field of view and mounting height. A wide triple-door entrance often needs more than one unit, spaced so their fields of view abut without gaps. Costs scale with the number of units, but the count zone is continuous: anyone crossing the line is counted once, regardless of which door leaf they walked through.
Door-frame on multiple leaves
Door-frame counters scale by leaf. A double-leaf entrance needs two units, a triple needs three. That is sometimes simpler to specify, but it produces a different kind of error: the boundary between two leaves becomes ambiguous. A visitor who drifts diagonally across the entrance, entering near the left leaf and exiting near the right, can be read as two events on separate counters, or as one event by neither. Counts have to be deduplicated downstream, which moves the accuracy problem into the analytics layer.
Wiring cost versus miss cost
On install cost, door-frame counters often look cheaper on paper. A unit at chest height needs a short cable run to a nearby socket, and a fit-out crew can usually mount one without scaffolding or a cherry picker. A ceiling-mount over a three-metre soffit needs a working-at-height permit, a longer cable pull through the ceiling, and frequently a Power over Ethernet drop from a comms cupboard rather than a wall socket.
It is fair to say that the up-front install bill for a ceiling-mount counter is meaningfully higher than for a door-frame counter on most entrances. The exact gap depends on the building: a shopfront with an exposed services raceway above the door is cheap to wire; a listed building or a finished ceiling with no nearby data run is not. Treat any blanket figure with suspicion. Get a quote on your own building.
The cost the install bill does not show is the cost of misses. Suppose a busy weekend entrance moves 8,000 visitors a day and a door-frame counter misses 4 percent of them through occlusion (a plausible range, not a guarantee, and entirely dependent on the entrance). That is 320 missed visits a day. Over a year, on a single entrance, it is six figures' worth of count error. If that count feeds conversion rate, staffing models, lease negotiations, or capacity reporting, the install saving disappears against the cost of running the business on the wrong numbers. A more accurate count is sometimes the cheaper system over a five-year horizon, even when its install bill is higher on day one.
The honest framing is a trade-off, not a verdict: door-frame is cheaper to put in, ceiling-mount is more accurate where people enter as groups. For a quiet entrance, the cheaper install may win on total cost. For a busy or group-heavy entrance, the accurate counter usually does.
Retrofit considerations
Most counter installs are retrofits into a building that already exists, often a shop fit-out that has to happen overnight. The retrofit story is different for each mount.
- Ceiling-mount retrofit. Needs ceiling access, a route for cable from the device to a network point, and a power source. In a typical retail store with a suspended ceiling, that is usually achievable in a short evening visit. In a heritage building, a glass canopy, or a polished concrete ceiling, it is not, and a different mount may be the only realistic option. Confirm the ceiling type and the nearest comms cabinet before specifying.
- Door-frame retrofit. Easier on the building. The unit fixes to the existing door jamb and can usually share a nearby wall socket. The cosmetic question is whether a chest-height device on the doorway is acceptable in the brand environment. In a premium retail or hospitality setting, a small device on the soffit out of eyeline is often preferred over a more visible one at the doorway.
- Network and power. Both mount types benefit from Power over Ethernet because it collapses power and data into one cable and runs cleanly through ceiling voids. Where PoE is impractical, a battery-powered door-frame unit can fit, with the maintenance overhead that batteries imply. The decision lives alongside a wider PoE versus battery trade-off that deserves its own conversation.
- Aesthetic and approval. Landlords, listed-building consents, and corporate brand teams all weigh in on what can sit on a customer-facing facade. A flush ceiling-mounted device is often more acceptable than a visible door-frame unit. Confirm with the facilities and design teams before ordering hardware.
Practical procurement guidance
If you are specifying counters across a portfolio of sites, the decision is rarely one mount for everything. It is a decision rule applied per entrance. A useful starting frame:
- Default to ceiling-mount for any high-volume entrance. If the doorway routinely sees groups, prams, families, or shoulder-to-shoulder traffic, top-down geometry pays for itself on count accuracy. That covers most mall entrances, transit hubs, busy flagship stores, and large public buildings.
- Consider door-frame on low-volume, single-file entrances. Where visitors enter one at a time with clear gaps, occlusion is rare and the side-on read is accurate. Side doors, back-of-house entries, and small specialist stores often fall into this category.
- Check the ceiling before you commit. A heritage facade, a glass canopy, or a concrete soffit can rule out ceiling-mount on a specific entrance regardless of preference. Get a site survey rather than assuming.
- Specify per-zone coverage, not per-door. On a wide entrance, ask the vendor how many devices are needed to cover the count zone without gaps, and how the system deduplicates a visitor crossing more than one field of view. A clear, single answer per entrance is what you are buying.
- Ask what happens on the busiest hour. A sensor that reads cleanly at 200 visitors an hour can struggle at 2,000. Request accuracy figures specifically for the entrance's peak conditions, not for a calm test environment.
- Mind the privacy posture. Top-down depth sensing reads geometry rather than images, which keeps the system out of personal data. Side-mounted devices that derive silhouettes from a small camera need a closer look at what is captured and what is stored.
How Ariadne mounts and counts
Ariadne measures this with Hybrid Fusion, its patented camera-free method. Time-of-Flight depth sensing counts every visitor at the entrances, capturing geometry rather than images, while patented phone signal sensing follows movement through the interior, detecting the signals a phone emits even in airplane mode. The sensor streams both feeds to Ariadne, where Hybrid Fusion combines them into one trajectory per visit and computes counts, dwell, and paths. The streams carry no identifier: no MAC address, no device ID, no biometric data, and no camera is involved. Identifiers are stored only when a visitor explicitly opts in, which keeps the method GDPR-friendly and outside biometric territory.
In practice, the entrance count comes from a Time-of-Flight depth sensor mounted on the ceiling above the doorway, looking down on the threshold. One device covers the typical retail or office entrance from a three-to-four-metre soffit; wider entrances are covered by additional devices laid out so their fields of view abut. The same sensor architecture sits in transit, cultural, and city installs. The hardware lineup, including ceiling-mount Time-of-Flight devices and the supporting interior signal sensors, is set out in the Ariadne sensor lineup, and the measurement method is described in detail on the how-it-works page.
FAQ
Are ceiling-mount counters always more accurate than door-frame ones?
Not always, but usually on busy entrances. The advantage of ceiling-mount is top-down depth geometry, which separates people who enter as groups and resists occlusion. On a quiet entrance where visitors arrive one at a time with clear gaps, a door-frame counter can be just as accurate. The accuracy gap opens on busy doors with groups, prams, and side-by-side traffic, which is where most miss-count cost lives.
What is the typical install cost difference?
Door-frame counters generally cost less to install because they sit at chest height and need a shorter cable run with no working-at-height kit. Ceiling-mount typically needs a longer cable pull, a Power over Ethernet drop, and access at three to four metres, which adds labour. Specific figures depend heavily on the building and on whether there is an existing comms point near the entrance; treat any blanket cost claim with caution and get a per-site quote.
Does Ariadne use a camera at the entrance?
No. Ariadne counts with Hybrid Fusion: Time-of-Flight depth sensing plus patented phone signal sensing, never cameras. Time-of-Flight captures geometry rather than images, and signal sensing captures no MAC address by default, so the measurement involves no video, no faces, and no biometric data.
How many ceiling-mount devices do I need for a wide multi-door entrance?
Enough to cover the full width of the count zone without gaps. A single Time-of-Flight unit at three to four metres typically reads a count zone roughly two and a half to four metres wide, depending on the sensor's field of view. A wide entrance with three door leaves often needs more than one unit. Confirm the count zone width and mounting height with the vendor against your specific entrance before ordering hardware.
